The Application of the Drop Volume Technique to Measurements of the Adsorption of Proteins at Interfaces EVA TORNBERG Chemical Center, Division o f Technology, University o f Lund, S-220 07 Lund 7, Sweden Received April 13, 1977; accepted July 15, 1977 A new procedure for the application of the drop volume technique to measurements of the rate of adsorption of proteins at interfaces has been developed. The mode of adsorption of the proteins lysozyme, fl-lactogtobulin and bovine serum albumin (BSA) at the air-water interface has been measured with the drop volume method and has been compared to measurements with the Wilhelmy plate technique. Due to surface enlargement of the drop throughout the process of the surface tension decay, slower kinetics of the adsorption process is obtained by the drop volume method compared to the Wilhelmy plate technique, and the proteins investigated were differently sensitive to this surface expansion. The adsorption process of the proteins has been evaluated in terms of different rate-determining processes. Different intermediate states between the native and the denatured forms have been observed. INTRODUCTION
The interfacial tension (y) is an important factor in the formation of emulsions and foams, although it has been pointed out by Van den Tempel (1) that it is not so much the absolute value of 3' which is crucial, but rather the gradients in 3' that arise during the emulsification or foaming process. The choice of the appropriate technique to measure the interfacial tension depends on the relationship between the phenomenon to be investigated and the information that the method can give. In the specific case of proteins stabilizing foams and emulsions, the kinetics of adsorption are of particular interest, and thus techniques which give dynamic rather than equilibrium measurements of y, give results which are more relevant to emulsification and foaming. There are few comparisons of available techniques for interracial tension determinations reported in the literature, especially for the case of time-dependent interfacial tension (2-5). The ring method is unsatisfactory for time-dependent solutions, as it can give errors as large as 10 mN m -1 (2), 391 Journal of Colloid and Interface Science, Vol. 64, No. 3, May 1978
and one main source of error is probably the act of detachment of the ring during measurement (3). The heights obtained in the capillary-rise method are found to be affected by aging of the surfaces over considerable periods of time (3). Both Boucher et al. (4) and Addison and Hutchinson (5) have compared the Wilhelmy plate, the drop volume and the pendant drop methods to measure the surface tension decay of surfactant solutions. Boucher et al. (4) found the Wilhelmy plate, the drop volume and the pendent drop methods to give almost the same values of the equilibrium surface tensions of surfactant solutions. On the other hand Addison and Hutchinson (5) found a difference between the rate of adsorption and the equilibrium values obtained by the pendent drop and the Wilhelmy plate technique. Lower y values were found by the latter technique. The y-t-curves obtained with the drop volume and the pendent drop methods coincide, which is reasonable since these methods are virtually the same, as discussed elsewhere (6, 7). The aim of this article is to demonstrate the applicability of the drop volume 0021-9797/78/0643-0391502.00/0 Copyright© 1978by AcademicPress, Inc. All rightsof reproductionin any formreserved.
392
EVA TORNBERG
technique to the elucidation of the mode of adsorption of proteins, and to compare it to the Wilhelmy plate technique. The following procedure was used. A drop of a certain volume, corresponding to a certain y-value, is expelled rapidly, and the time necessary for the surface tension to fall to such a value that the drop becomes detached is measured. Due t o the surface tension decay from the time of formation until detachment of the drop, the surface of the drop expands. The adsorption behavior of a protein can, in this case, be considered to be similar to adsorption at a newly created surface in a foam or an emulsion, with no further disruption of the interface. In this p a p e r , only surface tension measurements at the air-water interface are given, and they are compared to surface tension data obtained by the Wilhelmy plate method. The proteins lysozyme, fi-lactoglobulin and bovine serum albumin (BSA) were chosen for this study because their surface behavior from solution has already been subject to considerable study (8-17). EXPERIMENTAL
METHODS
Materials The method used for water purification is described in detail elsewhere (6). Three times recrystallized egg-white lysozyme, once recrystallized BSA and fl-lactoglobulin (A and B) from Sigma Chemical Co. were all lyophilized, and used without further purification. The three proteins were dissolved in a phosphate buffer (pH 7 and ionic strength, 0.1) as described elsewhere (16). The protein concentration used was 10-e% (w/w). In the investigation of the concentration dependence of the surface tension of lysozyme solutions, the solutions were prepared by dissolving lysozyme in 1 M KC1. For concentrations less than 0.1% (w/w), the solutions were prepared by diluting the Journal of Colloid and Interface Science, Vol. 64, No. 3, May 1978
stock solution of 0.1%. Acidity was adjusted to pH 3 with HC1. All protein solutions were freshly made each day.
Methods Density measurements of the solutions were carried out as described earlier (6). The description of the surface tension apparatus and its mode of operation according to the drop volume principle are given elsewhere (6).
Outline of Procedure The procedure for making measurements with pure liquids and solutions t t a t come to equilibrium quickly has been described earlier (6). For measurements of timedependent surface tensions a modified procedure is necessary. Part of the total drop does not fall, and the remaining portion will have its surface tension altered by adsorption of a surface active material; Hence, this aged surface m u s t be eliminated when the next drop is to be formed. Thus, by forming 1 or 2 drops rapidly, the next drop to be measured should be free of aged surface. The rapidly formed drops should have a rate of drop formation such that the liquid flow down the tube does not influence the resulting volume of the drops. This technique is limited to slow adsorption processes, as in the case of protein adsorption, because the rate of adsorption of the surface active material at the interface should be negligible compared to the time of formation of the drops, which is approximately 20 sec at the air-water interface. Because different drop sizes give rise to residual volumes of differing quantities, the residual volume (vr) below the orifice of the rapidly formed drop must be calculated in order to know the volume of the following drop. This entails the calculation of the maximum volume (Vm) before the falling part (vd) breaks away, and the
393
PROTEIN ADSORPTION AT INTERFACES
measurement of the detached volume (va). Harkins and Brown (18) calculated vm of an "ideal" drop according to the relation vmApg = 2Iir% This relation is, however, not valid, because (i) it does not take into account the fact that surface tension forces do not always act at an angle of q ) = 9 0 ° to the tip, and (ii) it also neglects the downward force due to the excess pressure inside the drop (7, 19). To avoid these errors the equation describing the volume of a pendent drop can instead be used to compute the maximum volume below the orifice just before it becomes detached. This has recently been done in reduced terms, Vm = Vm/a 3, by Boucher and Evans (7) as a function of X = r/a, where a 2 = 2y/Apg. Harkins and Brown (18) claimed that the volume of the falling fraction (va) of the maximum hanging drop (Vm) is determined by the shape of the drop, and this in turn depends solely on the value r/va ils or r/a. It may be argued that the break away process should be analyzed in terms of dynamics of flow during detachment before this assumption can be justified. Unfortunately, this analysis involves extremely complicated computational problems (21), and therefore Pearson and Whitaker (20) and Whitaker (21) have used another approach to avoid these problems. By means of photographic evidence Pearson and Whitaker (20) showed that the break away process for a 1.5 x 10-2 M heptanoic acid solution was essentially identical to that exhibited by water. Whitaker (21) has studied the stability of a cylindrical column of liquid as a model for the break away process, and the results obtained suggest that the drop stability is independent of the rheological properties of the surface, and that the geometry of the break away process depends only on v/r3; this is in accordance with the experimental findings of Harkins and Brown (18). On the basis of these recent findings the following procedure has been worked
out. The detached volume va of a quickly formed drop is measured, and from this volume the dimensionless quantity r/va 1/3 is calculated. For a given liquid there is only one unique value of r/a = X for each value of r/va 11~, and Wilkinson (22) has derived a fourth-power polynomial relationship between these quantities, which is used in this study. Using the tables given by Boucher and Evans (7) the value of X corresponding to r/va ~/~ gives the maximum pendent drop volume in reduced terms, Vm = vm/a 3. The residual volume (v~) can then be calculated according to v~ = vm - r e . The next drop to be measured is extruded by adding a certain amount of liquid, vs. The total amount of this new drop hanging beneath the tip is then obtained from the sum of v~ and yr. A new dimensionless quantity r3/vm is formed, and this can give the corresponding value of X by use of the tables of Boucher and Evans (7). The surface tension can then be determined from the equation for the capillary constant, a: y -
Apgr 2 2X 2
[1]
The comparison between the two methods of determining the interracial tension was performed at 25°C, whereas the measurements on the lysozyme solutions at different protein concentrations were made at 20°C. Theory A p p l i e d to Protein Adsorption
It is reasonable to assume that any process responsible for the time dependence of the reduction in surface tension by a polymer molecule must involve an increase in the number of adsorbed segments per unit area with time (23-25). In the case of protein molecules adsorbing at interfaces, Graham (14, 26) has shown that the primary layer of molecules is largely responsible for determining II, and the influence on II of an increase in Journal of Colloid and Interface Science, Vol. 64, No. 3, May 1978
394
EVA TORNBERG
the film thickness during adsorption can therefore be assumed to be negligible. The rate of reduction of interfacial tension is determined by three consecutive or concurrent processes (14, 27): (a) the diffusion of whole protein molecules to and attachment at the interface; (b) spreading or unfolding of already adsorbed molecules; (c) molecular rearrangements of adsorbed molecules. The two latter mechanisms involve transport of molecular elements or protein segments on the surface. A useful analysis of the kinetics of adsorption derives from the work of Ward and Tordai (28, 29), who considered the effects of diffusion from the bulk liquid to the surface and the energy barrier which the molecule must overcome in order to be adsorbed. They were able to show that this mechanism could describe the kinetics of adsorption of fatty acids at the hexane/water interface. In the absence of convection and desorption, the number of molecules per unit area (n) adsorbed at time t, providing the process is diffusion-controlled, is given according to Ward and Tordai (28) by: /
Dt
~1/2
The quantity D is termed the diffusion coefficient. It will only be equivalent to the conventional diffusion coefficient if there is no energy barrier between the surface and the subphase, and this is the case only at the very beginning of the process. In the general case, the value of D will be a quantity characteristic of both diffusion and the crossing of the barrier. In Eq. [2] it is assumed that the diffusion is an ideal process, that is to say solute-solute interaction is negligible. In the very dilute systems considered here, this approximation would be justified. In the case of protein adsorption the reduction of interfacial tension will be related to the number of active groups of Journal o f Colloid and Interface Science, Vol. 64, No. 3, May 1978
protein segments adsorbed. If v is the number of active groups per molecule, uCpr should be used for the concentration of active groups (ag). By analogy, the number of molecules per unit area, n, will instead be nag , i.e., the number of adsorbed active groups per unit area. When the modified form of Eq. [2] is combined with the simple two-dimensional equation of s t a t e H/nag = k T / v , which is considered to be obeyed in the relatively low pressure region, where diffusion is the rate-determining step (30), we obtain: Y = To - 2CprkT
t
[3]
A plot of y against t 1/z will then be linear. When a sufficient high energy barrier exists, the diffusion will no longer be rate-determining, but the rate of segment, or protein, penetration into the surface film will be rate-limiting. Ward and Tordai (29) considered this barrier to consist of the work IIAA, done in creating a space of area hA in a surface film of surface pressure II in order to adsorb a molecule. They formulated a differential equation for the kinetics of adsorption, and if we assume irreversible adsorption the equation is: dn
-- klCpr e x p [ - I I A A / k T ] .
[4]
dt
In this equation n is the concentration of the solute adsorbed (number of molecules per unit area), and kl is the rate constant of adsorption. MacRitchie and Alexander (8) have used this equation for the adsorption of protein molecules from the bulk to the interface. This work has been commented on by Bull (31), who considered it more likely that the decay of the surface tension with time is not entirely due to the adsorption of new protein molecules to the surface, but also involves the expansion and rearrangement of the protein molecules on the surface.
395
PROTEIN ADSORPTION AT INTERFACES
•
'E
70
•
Drop volume
x Withelmy
Lysozyme
prate
--
- ?.-tactogtobu~in
--BSA
g
...........
i iiiii,i iii
5010
20
3
30
40
Time ( m i n u t e s ]
Fro. 1. Time-dependence of interfacial tensions at the air/water interface for 10-2% (w/w) solutions of Iysozyme, fl-lactoglobulin and BSA measured with both the drop volume and the Wilhelmy plate methods.
To take into account the various processes than can occur during the surface tension decay, Eq. [4] can be modified. By assuming that y reflects the level of adsorption of protein segments and/or active groups (ag), dnag/dt can be considered to be proportional to dII/dt. The effective concentration of ag is now given by vCvr. Equation [4] will consequently be modified to give Eq. [5]: In --dH = In (klb, Cpr) - HAAag/kT.
dt
[5]
If z~tAag is assumed to be constant, then a plot of In (dII/dt) vs II should be linear with a slope giving AAag, which will represent the mean area created in the film in order to adsorb an active group in the protein molecule. The values of AAag , iv, ]~1 and Cpr can be considered to be constants within each rate-determining process, but these values will change when the rate-determining process is changed. When the rate is determined by diffusion (Eq. [3J) In (dII/dt) is not formally a linear function of 1S as in Eq. [5J. We find, however, for all practical purposes and for the range of H-values of interest here, that In (dII/dt) can be represented by Eq. [5]. When in (dH/dt) is independent of U, i,e,, when aA~g ~ 0, this is taken as an indication of only rear-
rangement processes taking place. No new areas, AAag, a r e created in the film, but the surface coverage is still enhanced by protein segments adsorbing at the surface unoccupied by protein segments in the preceding processes. Equation [5] is expected to be valid as long as the effect of collapse or coagulation is negligible. RESULTS AND DISCUSSION
In Fig. 1 the interracial tension has been plotted as a function of time up to 40 rain for the three proteins investigated, using both the drop volume and the Wilhelmy plate method. The initial interracial tension (Y0) equal to the interracial tension of the buffer solution, was measured to be 72.4_+ 0.4 mN m -1. In the case of the three proteins studied here, the drop volume method indicated slower kinetics than obtained with the Wilhelmy plate, which is in accordance with an enlargement of the surface of the drop throughout the process of the surface tension decay. Because of the large flat surface in the Wilhelmy plate method, eddy and convective currents more frequently occur than in the drop volume method, which can also contribute to the differences in the y-t curves obtained. It is interesting to compare the surface behavior of these proteins when the surface Journal o f Colloid and Interface Science, V o | . 6 4 , N o . 3, M a y 1978
396
EVA TORNBERG i'\ f ~
.~"
o..,o e°'° to"
*'
i !x
_~
I I=, ~
x
'%~
• Drop vo ume x Wilhe[my plate
...... Lysozyrne --tl-iactoglobulin
,,,
"o "'X~ " ,e.e,.
Time112 (minutes 1/2 )
FIG.2. Interracialtensions at the air/waterinterfaceas a functionof timet/2for 10-2%(w/w)solutions of lysozyme,/3-1actoglobulinand BSA measured with both the drop volumeand the Wilhelmyplate methods. is enlarged. As can be seen from Fig. 1, the y-t curve of fl-lactoglobulin is not as sensitive to the surface expansion, whereas the rate of the surface tension depression by BSA and lysozyme is greatly diminished by the expanding surface in the drop volume method. In Fig. 2 y, read off from the fitted curve in Fig. 1, has been plotted against t 1/~ for lysozyme,/3-1actoglobulin and BSA according to Eq. [3]. There is an initial period before a linear relationship is established for the measurements with the drop volume method, and this is especially pronounced for lysozyme. This phenomenon can probably be attributed to the enlargement of the surface of the drop, accomplished by the surface tension decay of the first arrived molecules, being faster, than the process of building up an activation barrier against further adsorption, which is also a consequence of increasing II. Therefore, the rate of arrival of protein segments or molecules at the surface, as reflected in dII/dt, increases during this initial period as a function of H. This is shown by the results for Journal of Colloid and Interface Science, Vol. 64, No. 3, May 1978
lysozyme given in Fig. 3, where log (dII/dt) is plotted as a function of II. The maximum slope, which corresponds to the linear part of the plots marked with two arrows given in Fig. 2 is always taken for quantitative interpretations. It is assumed that the supply of protein molecules by diffusion is the limiting factor during this stage. After the linear part of the y - t 1/2 curve, the rate of decrease in surface tension with increasing t 1/~ falls off. In Fig. 3, log (dII/dt) has been plotted against H for the three proteins, measured with the Wilhelmy plate and the drop volume methods. The slopes were obtained by computer differentiation of the fitted curves of Fig. 1, which were approximated by parabolas for small sections. Linear parts derived by linear regression analysis were obtained with evident breaking points for all the protein solutions. The correlation coefficients for the linear parts obtained could vary from 0.70 to 0.99. The linear behavior in the plots indicates that the theory on which Eq. [5] is based, is not invalid, and that the rate of desorption is negligible. The
PROTEIN ADSORPTION AT INTERFACES • Drop
~-~~i~'-o \ •
volume
x Wilhe{my
..........
plate
\
397 Lysozyme
.....
I~-[actoglobu[in
-
BSA
-
\
\,\ • "
..
":~,.
"~,
• •
~lz,° o~
•
"~.x\ i ,\
-1
rr
( r a N m "1 )
Fro. 3. Log (dIi/dt) as a function of II for I0-2% (w/w) solutions of lysozyme,/3-1actoglobulinand BSA adsorbing at the air/water interface and measured with both the drop volume and the Wilhelmy plate methods.
graphs of Fig. 3 show that the surface behavior of proteins has a step character, more or less obvious, and that in some cases the abrupt change from one step to the consecutive one is highly indicative of a cooperative process taking place at the interface. To summarize the quantitative evaluation of the graphs from Figs. 2 and 3, the surface pressure (Ilatt) and AAag attained during the different rate-determining steps are collected in Table I. AAag was evaluated from the slopes of all the linear parts in Fig. 3 according to Eq. [5]. Atatt/40 and AI][att/II40 a r e quotients, which indicate, respectively, the relative contribution of the time elapsed and Ilat t o f each rate-determining step in relation to 40 rain and to the surface pressure attained after 40 min. Surface pressures attained during t h e " d i f fusion" steps in Table I, denoted as Iltl~2, evaluated from the curves in Fig. 2, and the surface pressures for the first steps derived from Fig. 3, denoted as HA, are the same within the limits of error, which implies that it is possible to distinguish when diffusion or penetration of the surface film is the rate-
controlling factor. Unfortunately, the diffusion step is too fast to be detected with the Wilhelmy plate method (see Fig. 2), so these values are only estimates. The surface pressures attained during the initial step, when the effect of surface enlargement is evident, also coincide well, when evaluated from Figs. 2 and 3 for the interfacial adsorption of lysozyme. Diffusion controlled adsorption of proteins at an interface can imply either that the molecules diffuse to the interface and adsorb without further spreading, or that spreading or unfolding is so quickly performed that diffusion becomes the ratecontrolling factor. The penetration of a surface film can be accomplished either by spreading or unfolding of already adsorbed molecules or by adsorption of additional molecules arriving at the interface. But in order to adsorb a protein molecule or segment at the surface, a surface a r e a , AAag , has to be cleared in the film, which demands molecular rearrangements of the already adsorbed protein molecules. Graham (14) has been able to show the existence of two kinetic regions during the penetraJournal of Colloid and Interface Science, V o l . 64, N o . 3, M a y 1978
398
EVA TORNBERG TABLE I
Parameters Describing the Kinetics of Surface Tension Decay of Lysozyme,/3-Lactoglobulin and BSA Solutions of 10-2% (w/w) at the Air/Water Interface Measured with Both the Drop Volume and the Wilhelmy Plate Methods I]att
Protein
Lysozyme
fl-Lactoglobulin
Method of measurement
Drop volume
AA per ag (nm2)
AI~att/II40 (%)
Atatt/40 (%)
--
17
18
2.0
Diffusion Penetration
11.2 --
10.5 11.8
0.2 2.6
72 I1
58 24
Wilhelmy plate
Diffusion Penetration Penetration
8.4 ---
8.4 11.3 14.4
-2.4 3.4
58 20 22
1 10 89
Drop volume
Diffusion Penetration Penetration Rearrangement
12.8
11.0 17.3 18.0 20.0
0.4 2.5 10.6 --
55 32 3 10
1 12 14 73
Diffusion Penetration
13.6
13.4 18.0 19.6
-2.5 6.7
68 24 8
0.3 12.7 87
10.8 16.3
0.9 2.0
66 34
10 90
--
16.9 18.8
-5.4
86 10
8
--
19.6
9.5
4
91.5
Drop volume Wilhelmy plate
Effect of surface enlargement
IlttJz IIa (raN rn -~)
2.1
Wilhelmy plate
BSA
Rate-determining step
---
--
--
Penetration
--
Diffusion Penetration
--
Diffusion Penetration Penetration
tion-controlled adsorption of/3-casein, BSA and !ysozyme, the first one describing the adsorption when additional molecules can arrive at the interface, the second one involves only rearrangements within the film with no further adsorption of m o l e c u l e s from the bulk, Although the results presented in Table I have only been recorded up to 40 min, two discrete penetration steps with discernible activation barriers can be observed for the proteins studied, anyhow, when measuring with the Wilhelmy plate method. Comparing the two penetration steps, the first one is always faster and demands a smaller area, AAag , to be swept out in the film, whereas the second one is slow and the rearrangements o c c u r with a greater n u m b e r of residues, i.e., with a larger AAag. This is conJournal of Colloid and Interface Science, Vol. 64, No. 3, May 1978
8.1
16.4
0.5
sistent with the larger the surface area, h A a g , to be cleared, the more n u m b e r of residues to be rearranged within the surface film, and the slower the process is. Graham (14) has shown that rearrangements occur with a greater n u m b e r o f segments in a protein film with greater structure and order. This is in accordance with hAag increasing with IL The last step occurring for fl-lactoglobulin, shown~in Fig. 3, when using the drop volume method, gives log (dII/dt) independent of II, which is taken to indicate the occurrence of rearrangement processes taking place, with no need of new areas to be swept out in the interface. F u r t h e r support for this assumption is that this rearrangement step follows directly after a penetration step~ with a high value of AAag.
399
PROTEIN ADSORPTION AT INTERFACES
The supply to the trolled
results in Table I indicate that the of protein molecules or segments interface is mainly diffusion conin all cases (compare the values of z~l[Iatt/II40), but the rate of diffusion differs considerably due to the protein adsorbed. Comparing the values of the quotient Atatt/ 40, the diffusion of/3-1actoglobulin is most rapid followed by BSA and lysozyme, the latter having a comparatively slow diffusion. Evidently, a slow diffusion controlled adsorption, as in the case of lysozyme, seems to be a condition for the effect of surface enlargement to become evident. The slow diffusion of lysozyme in comparison with the two other proteins seems to be contradictory, as the molecular weight oflysozyme (-~ 14,600) is smaller or in the same order of magnitude as for /3-1actoglobulin (-~18,500), BSA having the highest molecular weight (~66,000). Considering not only the diffusion coefficient of the molecules, but also the probability of successful collision with the interface resulting in adsorption, as variables governing the diffusion-controlled adsorption, electrostatic repulsion between adsorbing species and the hydrophobicity of the molecules (14) have to be taken into account. As lysozyme among the proteins studied is furthest away from the isoelectric point, the net charge of the molecule is highest;
~"
hence the electrostatic hindrance during adsorption will probably contribute to a diminished number of successful collisions, being one of the reasons why diffusion of lysozyme is slowest of all the proteins studied. Evidently the probability of successful collision of molecules with the interface is of great importance in making the diffusioncontrolled process fast, which is also the experience of Graham (14). Comparing the two methods of measurements, the first to be noted is that the effect of surface enlargement during the initial step of diffusion does not occur with the Wilhelmy plate method. The next is that the diffusion-controlled step mostly lasts to a higher surface pressure, H, and is much faster, when using the Wilhelm~, plate than the drop volume method. This is especially obvious for BSA. Probably due to the surface renewal during the surface tension decay in the drop volume method, the diffusion step is slowed down, and the unfolding step is promoted. Both these phenomena enhance the possibility of concurrent unfolding or spreading of adsorbed molecules during the diffusion-controlled step, when measuring with the drop volume method, which will lead to a buildup of an adsorption barrier at an earlier stage of the adsorption process, i.e., at a lower II. This reasoning will hold as long as unfolding or
7C
O.OOI
%
o.T
°/°
1,0
%
E
g -tl
fig
"~'~-L-~-JL--~
,
o o
1~
2'o -
-~
3'0
~o
Time { m i n u t e s )
FIG. 4. Time-dependence of interfacial tensions at the air/water interface for 1 M KC1 solutions of lysozyme from 1.0 to 0.001% (w/w). Journal o f Colloid and Interface Science, Vol. 64, No. 3, M a y 1978
400
EVA TORNBERG
• 0.001% 75
= '¢"
;
4
,,t
~
g
"~'l~h~
• ~
z
I~01 %
x 0.1
%
o 1.0
%
70
_~65 -~ 6O
,T Time 1/2
(minutes 1/2)
FIG. 5. Interracial tensions at the air/water interface as a function of time l/2 for 1 M KC1 solutions of l y s o z y m e from 1.0 to 0.001% (w/w).
spreading of molecules gives rise to lowered compressibility of the surface film. For lysozyme the diffusion step does not last to a lower II with the drop volume method, which can be attributed to its low capability to unfold at an interface, which has been demonstrated by Graham (14). It is also interesting to note that the diffusion of /3-1actoglobulin is comparatively rapid and
is not much slowed down, when measuring with the drop volume method. This could be one of the reasons, why the y - t curve obtained with the drop volume method is not so much altered compared to the curve obtained with the Wilhelmy plate method, as seen in Fig. 1. To illustrate the concentration dependence of 1 M KCI solutions of lysozyme,
o 1.0" % 0.1
x~
g,
x
• 001
%
%
=¢
0
o oo •
0
A,I~
•••
K
o
e
1
i
5
I0
15
20 •rr
z~
(rnN m-t )
FIG. 6. Log (dFlldt) as a function of H for 1 M KCI solutions of l y s o z y m e from 1.0 to 0.001% (w/w) adsorbing at the air/water interface. Journal of Colloid and Interface Science, Vol. 64, No. 3, May 1978
401
P R O T E I N A D S O R P T I O N AT I N T E R F A C E S T A B L E II
P a r a m e t e r s Describ ing the Kinetics of the L o w e r i n g of Surface Te ns i on of 1 M KCI Solutions of L y s o z y m e at the A i r - W a t e r Interface M e a s u r e d with the Drop V ol ume M e t h o d Ilatt Rate-deter-
Concn % (w/w)
HA
~/1/2
mining step
(mN m ~)
AA per ag (nm2)
AglatjII40 (%)
Atatt/40 (%)
1
Diffusion Pen etration Pen etration
13.6 ---
12.6 19.1 21.0
0.5 2.1 7.3
60 31 9
2 17 81
0.1
Diffusion Penetration Penetration
8.6 ---
6.7 15.3 18.0
0.4 1.0 5.2
37 48 15
7 30 63
0.01
Diffusion Penetration
10.3 --
7.8 11.5
0.2 0.7
68 32
65 35
the surface tensions at the a i r - w a t e r interface derived by the drop volume method are given as a function of time up to 40 rain in Fig. 4. The most striking feature to be seen in Fig. 4 is a faster change in, as well as a lowering of, the surface tension with increasing bulk concentration. Lankveld and Lyklema (23) have also shown that the rate of decrease in interfacial tension is faster the higher the polymer concentration, for poly(vinyl alcohol) adsorbed at the paraffin-oil/water interface. Adams et al. (10) deduced from their results with lysozyme films that low bulk concentration favors the formation of an adsorbed film containing a mixture of unfolded and native molecules, while a high concentration leads to the formation of a film of mostly native molecules. In analogy with the earlier evaluation of the kinetics, ~ - - I 1/2 c u r v e s and log ( d H / d t ) II curves are presented in Figs. 5 and 6, respectively, for all the concentrations. In Table II the parameters evaluated from the curves (cf. to Table I) are given for the different concentrations. From the curves in Figs. 5 and 6 it may be concluded that the effect of surface enlargement during the diffusion step becomes more evident, as the bulk protein concentration diminishes, being visible in Fig. 6 for concentrations -<0.01% (w/w). When
diffusion as a rate-determining step persists over longer times at lower concentrations, as seen for the quotient Atatt/40 in Table II, the probability of the surface enlargement effect increases, as also suggested from the results in Table I. At the highest concentration of 1% (w/w) the step of penetration does not start until a surface pressure of about 13 mN m -I, whereas at the lower concentrations the diffusion step lasts to a surface pressure of about 8 mN m -1. This is in accordance with the lysozyme molecules spreading more easily at lower bulk concentrations during the diffusion step. The enhanced diffusion controlled adsorption to the interface during the first 40 min by the concentration of 0.01% as compared to concentration of 0.1% can be due to the low flexibility of the lysozyme molecules at the interface. Compared at the same surface pressure, the value of zXAag increases with decreasing concentration, which in combination with the reduced diffusion rate implies that the rate of interfacial tension decay is slowed down at lower concentrations, which is also observed in Fig. 4. NOMENCLATURE
Cpr
protein concentration in molecules/ liter
Journal of Colloid and Interface Science, Vol. 64, No. 3, May 1978
402 kl
k~
n nag r /)d Um Vr
1.)s AA AAag
EVA TORNBERG
rate constant for the adsorption o f m o l e c u l e s at an interface modified rate constant for the adsorption o f active groups within a protein m o l e c u l e at an interface number o f adsorbed m o l e c u l e s per unit area number o f active groups adsorbed per unit area radius o f the tip
detached volume of a quickly formed drop maximum volume of a drop hanging b e l o w the orifice residual v o l u m e b e l o w the orifice after a rapidly formed drop has detached the v o l u m e extruded for the drop to be measured mean surface area created in a film in order to adsorb a m o l e c u l e mean surface area created in a film in order to adsorb an active group in the protein m o l e c u l e number o f active groups per protein molecule ACKNOWLEDGMENTS
Professor K. Larsson and Dr. David Graham are heartily thanked for valuable comments. The skillful technical assistance of Mrs. Gunnel Lundh is gratefully acknowledged. The author wishes to thank Mrs. Lisbeth Rydhag (The Swedish Institute for Surface Chemistry, Stockholm) who has kindly made the measurements with the Wilhelmy plate method. The investigations were made possible by grants from the Swedish Board for Technical Development, and this financial support is gratefully acknowledged. REFERENCES 1. Van den Tempel, M., Proc. 3rd Int. Congr. Surface Activity, Cologne 1960, 2, 573. 2. Padday, J. F., in "Surface and Colloid Science" (E. Matijevic, Ed.), Vol. 1, p. 39. Wiley (Interscience), New York, 1969. 3. Padday, J. F., and Russell, D. R., J. Colloid Sci. 15, 503 (1960). 4. Boucher, E. A., Grinchuk, T. M., and Zettlemoyer, A. L., J. Colloid Interface Sci. 23, 600 (1967). Journal of Colloidand Interface Science, Vol. 64. No. 3. May 1978
5. Addison, C. C., and Hutchinson, S. K., J. Chem. Soc. 3387 (1949). 6. Tornberg, E., J. Colloid Interface Sci. 60, 50 (1977). 7. Boucher, E. A., and Evans, M. J. B., Proc. Roy Soc. A 346, 349 (1975). 8. MacRitchie, F., and Alexander, A. E., J. Colloid Sci. 18, 453 (1963). 9. Hamaguchi, K., J. Biochem. 31, 123 (1958). 10. Adams, D. J., Evans, M. T. A., Mitchell, J. R., Phillips, M. C., and Rees, P. M., J. Polymer Sci. 34, 167 (1971). 11. Graham, D. E., and Phillips, M. C., in "Theory and Practice of Emulsion Technology" (A. L. Smith, Ed.), p. 57. (Symp. at Brunel Univ., 16-18 Sept., 1974) Academic Press, New York, 1976. 12. Phillips, M. C., Evans, M. T. A., Graham, D. E., and Oldani, D., Colloid Polymer Sci. 253, 424 (1975). 13. Yamashita, T., and Bull, H. B., J. Colloid Interface Sci. 24, 30 (1967). 14. Graham, D. E., thesis, Unilever Research Laboratory, Colworth/Welwyn, The Frythe, Welwyn, Hertfordshire, 1976. 15. Joos, P., Biochim. Biophys. Acta 375, 1 (1975). 16. Evans, M. T. A., Mitchell, J. R., Musselwhite, P. R., and Irons, L., Advan. Exp. Med. Biol. 7, 122 (1970). 17. Muramatsu, M., and Ishii, T., Bull Chem. Soc., Japan 43, 2364 (1970). 18. Harkins, W., and Brown, F. E., J. Amer. Chem. Soc. 41, 499 (1919). 19. Hartland, S., and Srinivasan, P. S., J. Colloid Interface Sci. 49, 318 (1974). 20. Pearson, F. W., and Whitaker, S., J. Colloid Interface Sci. 54, 219 (1976). 21. Whitaker, S., J. Colloid Interface Sci. 54, 231 (1976). 22. Wilkinson, M. C., J. Colloid Interface ScL 40, 14 (1972). 23. Lankveld, J. M. C., and Lyklema, J., J. Colloid Interface Sci. 41, 454 (1972). 24. Frish, H. L., and Simha, R., J. Chem. Phys. 24, 652 (1956). 25. Frish, H. L., and Simha, R., J. Chem. Phys. 27, 702 (1957). 26. Graham, D. E., personal communication. 27. B6hm, J. T. C., thesis, Meded. Landbouwhogeschool Wageningen 74-5, 1974. 28. Ward, A. F. H., and Tordai, L., J. Chem. Phys. 14, 453 (1946). 29. Ward, A. F. H., and Tordai, L., Recueil 71, 572 (1952). 30. Davies, J. T., Biochem. Biophys. Acta 11, 165 (1953). 31. Bull, H. B. ,J. Colloid lnterface Sci. 41,305 (1972).